Family Medicine

World Book of Family Medicine – European Edition 2015
Tor Anvik, MD, Ass. professor
tor.anvik@uit.no
45 – Is There Such a Thing as a Correct Diagnosis in Family
Medicine? Learning Points from the Use of Consensus in Validating Diagnoses
in Family Practice
Tor Anvik, MD, Ass. Professor
Department of Community
Medicine, Faculty of Health
Sciences, University of Tromsø,
Norway
Background
A correct diagnosis is the basis for therapy in clinical medicine. Research in validation
of medical diagnoses has however demonstrated disappointing results. Different
doctors have been shown to make different diagnoses when presented with the same
case history, and the same doctor makes different diagnoses when the same history is
presented several times.
The author wanted to study the ability of experienced family doctors to make a
diagnosis based upon a video recording of real life consultations in family practice,
and to test the impact of subsequent information about the long-term outcome as
well as a small group discussion on agreement or disagreement about correct
diagnosis.
Setting and Methods
244 consecutive patients over the age of 15 presenting for consultations with the
author in a group practice in northern Norway, were asked for informed consent to
video recording of their meeting with the doctor. Immediately following each
consultation the doctor decided whether the patient had presented a new problem
that had not been discussed before. The doctor made a note of the new diagnoses
that he had made by the end of the consultations. The patients were interviewed by
telephone 3-6 months later. Three years later the medical records of the patients
were scrutinised and the doctor made a final diagnosis.
Altogether, 69 video recorded consultations where the patients presented 90 new
problems were included. In 20 problems a specialist had made the final diagnosis
after referral. Among the remaining 70 problems, 20 were selected at random for
validation by experienced FDs. Four patients did not consent to their video being
shown to other FDs, which allowed the remaining 16 video recorded consultations
with 16 new problems to be analysed further.
Five experienced family doctors were invited and each of them met three times in
groups of three. The video recordings and the medical history of the patients were
presented according to a step-by-step procedure:
1. The doctors were given a short written summary of the patient's previous history
and presented with one new problem which they were expected to diagnose. They
then watched the video recording of the whole consultation. After the video, the
author answered clarifying questions and elaborated on his clinical findings during the
consultation. The author then left the room. Each doctor was asked individually to
write down the most likely diagnosis of this single problem (Diagnosis 1), without
138